Abstract
The most important reason for the lack of a German nationwide and standardised survey of soil organisms is probably the time-consuming and expensive identification of soil invertebrates. The present contribution should contribute to solving this problem. Earthworms and soil were sampled at 25 sites, the animals were identified morphologically and by community DNA (comDNA) and environmental DNA (eDNA) metabarcoding. The comparison of results showed that comDNA detected more species (3.6 on average) than eDNA (3.0) and morphological identification (2.8). In contrast, eDNA, on average, detected a similar number of species as morphological identification. However, some species appear to have a different probability of being detected by eDNA than others, depending on their abundance, behaviour, biology or body size. All three identification methods can differentiate between sites with different species composition, and the degree of separation can vary depending on the identification method. The relative proportion of eDNA reads shows potential as a surrogate of relative abundance/biomass for endogeic but not for anecic species. The overall aim of the 'MetaSOL' project (which the present contribution originated from) was to develop recommendations for efficient and routinely implementable monitoring of soil fauna. The results showed that genetic identification methods are suitable for earthworms. Before genetic identification methods can be introduced into official practice, key preconditions such as comprehensive, well-curated and quality-controlled DNA reference databases and method standardisation must be addressed. Robust indices of soil health based on soil organism data need to be developed. The inclusion of further groups in addition to earthworms should be examined.